GuillaumeLeclerc/gopt

Issues

• experiment running init functions on cuda too

  #30 opened 4 years ago by GuillaumeLeclerc
  0

• Optmize the allocation of the direction vector using numba intrinsics

  #29 opened 4 years ago by GuillaumeLeclerc
  0

• Change Problem API to allow for exhaustive coordinate descent

  #28 opened 4 years ago by GuillaumeLeclerc
  0

• Implement 2-opt strategy for TSP

  #7 opened 4 years ago by GuillaumeLeclerc
  0

• Implement a better solution init strategy for TSP

  #13 opened 4 years ago by GuillaumeLeclerc
  0

• Suggest the user to use `Compiler.inline = 'always'` for long running instances

  #25 opened 4 years ago by GuillaumeLeclerc
  1

• Return a nice Result object instead of a tuple

  #22 opened 4 years ago by GuillaumeLeclerc
  0

• Implement Simulated Annealing

  #5 opened 4 years ago by GuillaumeLeclerc
  2

• Implement cuda Runner

  #24 opened 4 years ago by GuillaumeLeclerc
  0

• Have a nice default value for num_cores in cpu_runner

  #23 opened 4 years ago by GuillaumeLeclerc
  0

• Document time of init

  #21 opened 4 years ago by GuillaumeLeclerc
  0

• Experiment intel SVML

  #20 opened 4 years ago by GuillaumeLeclerc
  1

• Experience with different threading layers

  #19 opened 4 years ago by GuillaumeLeclerc
  1

• optimizer.init should not have access to problem_data

  #14 opened 4 years ago by GuillaumeLeclerc
  0

• Let user select number of cores

  #18 opened 4 years ago by GuillaumeLeclerc
  0

• Use np.nan instead of -1 for unintialized loss arrays

  #17 opened 4 years ago by GuillaumeLeclerc
  0

• Implement multi-core Runner

  #15 opened 4 years ago by GuillaumeLeclerc
  0

• Write proper docstrings

  #12 opened 4 years ago by GuillaumeLeclerc
  1

• Implement Shuffler API

  #8 opened 4 years ago by GuillaumeLeclerc
  0

• Disable numpy multicore execution

  #11 opened 4 years ago by hediD
  0

• Implement basinhopping

  #9 opened 4 years ago by GuillaumeLeclerc
  0

• Check in the Runners that the type of problem data matches what the code was compiled for

  #10 opened 4 years ago by GuillaumeLeclerc
  0

• Implement k-opt for TSP

  #6 opened 4 years ago by GuillaumeLeclerc
  0

• Do not call the loss if a neighbor function returns a loss delta

  #4 opened 4 years ago by GuillaumeLeclerc
  0

• Optimize TSP so that the neighbor function returns a delta loss

  #1 opened 4 years ago by GuillaumeLeclerc
  0

• Write Readme

  #3 opened 4 years ago by GuillaumeLeclerc
  0

• Make TSP as a generic problem provided by the library

  #2 opened 4 years ago by GuillaumeLeclerc
  0